[Prospective multicenter cohort study on breast cancer screening using an automated breast ultrasound with remote reading].

Abstract

Objective: To construct a remote screening network for breast cancer based on automated breast ultrasound (ABUS) and explore the value of ABUS with remote reading for breast cancer screening. Methods: We constructed a remote breast cancer screening network including one remote reading center and 48 image-acquisition centers. We recruited women to participate in breast cancer screening at one of these image-acquisition centers from January 2021 to January 2023. The technicians collected the whole breast images using the ABUS. The images were then sent to the reading center through the PVBUS System and interpreted independently by two radiologists using the Breast Imaging Reporting and Data System (BI-RADS). BI-RADS categories 1 and 2 indicate negative screening results, and women diagnosed with these categories were recommended for annual breast ultrasound screening. BI-RADS categories 3, 4, and 5 indicate positive results. Women with BI-RADS category 3 lesions were recommended for follow-up examinations every 6 months using ABUS or handheld ultrasound, while those with BI-RADS 4 and 5 lesions were suggested to undergo pathological examinations. Results: In our study, we enrolled 10 344 women who completed the ABUS screening and were followed up for more than 12 months. After remote reading, 6 164 women were diagnosed with BI-RADS category 1 and 2 626 woman were within BI-RADS category 2. In contrast, 1 404 women were within BI-RADS category 3, a total of 135 women were within BI-RADS category 4, and 15 women were within BI-RADS category 5. The positive screening rate of ABUS was 15.0% (1 554/10 344). The ABUS with remote reading had a detection rate of 3.7/1 000 (38/10 344) for breast cancer screening, with a sensitivity of 97.4% (38/39) and a specificity of 85.3% (8 789/10 305). Among the 38 breast cancer cases detected, 92.1% (35/38) were invasive carcinomas, and 63.2% (24/38) were stage 0 or Ⅰ breast cancers. Conclusions: Breast cancer screening based on ABUS with remote reading provided an efficient and feasible solution to the problem of unevenly distributed medical resources and medical staff levels in various regions of China, enabling the decentralization of high-quality medical resources and improving the accessibility of high-quality screening services. It has provided an alternative for breast cancer screening in China.